Getting to specifics, most children do not need a bolus of insulin for a midmorning snack that contains 15 grams of carbohydrate. However, if a child who eats a midmorning snack consistently has high blood glucose at lunch, we program a basal rate increase for one hour before snacktime. For example, if the usual basal rate is 0.2 units/hour in the morning, we would make it 0.7 units/hour at 9 AM and set the basal rate back to 0.2 units/hour at 10 AM. By covering the snack this way, neither the child nor the teacher has to worry about a bolus dose. We do the same thing in school-age kids if coverage of a snack is necessary.
For young school-age children, the parent generally tells the school nurse or teacher what dose the child needs for lunch based on blood glucose level, food intake, and activity. The dose may be different if lunch is followed by gym or recess rather than an academic class. Most children, even three-year-olds, can program their own boluses with adult supervision.
For organized sports, the pump is usually removed. If the pump contains lispro, we recommend disconnecting for no longer than two hours without bolusing. If a child is at the beach, he may hook up and bolus every two hours or take a lunch shot of Regular insulin, which will usually last for four to five hours.
If a child lives at the beach in the summer, we can substitute Regular for lispro in the pump for the whole summer. It means a few days of re-regulating the child, but it isn’t usually a big problem and it allows him to go for four to five hours without having to hook up to the pump. (Note: If you currently use a pump, do not attempt to make this change on your own. Work with a health-care provider experienced in pump therapy.)
At the time of writing, the Yale Program has 250 children on the insulin pump — close to half of all the children we see. We initiate treatment with at least two children per week. The mean glycosylated hemoglobin (HbA1c) level for all children on pumps at our center is 7.2%. (The HbA1c test gives an indication of overall blood glucose control over the previous two to three months.) While a person who does not have diabetes generally has an HbA1c of 4.2% to 6.3% in our lab, research has shown that an HbA1c of 7.2% or lower in people with diabetes is associated with a much lower risk of long-term diabetes complications.
Our philosophy is that intensive management of blood glucose levels is essential right from diagnosis, so all of our children with diabetes are intensively managed regardless of their treatment method. The mean HbA1c for all of them is 7.7%. The 0.5% decrease — without hypoglycemia — seen in those using a pump is a significant improvement.
Our hope is that all pediatric diabetes programs will offer the pump option to their children and families. To quote many of the parents we see, “It gave us our life back.” What an inspiration for us to continue our pump program!